Skip to main content

Step 5: Export to S3

The final step of this tutorial sends your enriched, restructured, and batched logs to Amazon S3.

The Goal​

You will replace the local file output in your pipeline with an aws_s3 output, configuring it to send your batched logs to an S3 bucket.

The aws_s3 Output​

The aws_s3 output processor is very similar to the file output, but requires a few extra fields for the bucket name, region, and AWS credentials.

Implementation​

  1. Start with the Previous Pipeline: Copy the batched-output.yaml from Step 4 to a new file named export-s3.yaml.

    cp batched-output.yaml export-s3.yaml

  2. Replace the Output with S3: Open export-s3.yaml and replace the entire output section with the aws_s3 block below.

    Replace the 'output' section in export-s3.yaml
    output:
      aws_s3:
        bucket: ${S3_BUCKET_NAME} # From your setup step
        region: ${AWS_REGION}
        # This path creates a simple date-based folder structure
        path: 'logs/${!timestamp_unix_date()}/${!uuid_v4()}.jsonl'

        # We can re-use the same batching policy from the previous step
        batching:
          count: 10
          period: 5s

        # You must configure your credentials. Using a profile is a common method.
        credentials:
          profile: 'default' # Or your named AWS profile

  3. Configure AWS Credentials: Ensure your environment is configured with AWS credentials. The pipeline above is configured to use the default profile. Make sure it's configured or change it to your named profile.

    # Check your default profile
    aws configure list

    # Don't forget to set the bucket name and region environment variables
    export S3_BUCKET_NAME="your-unique-bucket-name"
    export AWS_REGION="us-east-1"

  4. Deploy and Test: Run the pipeline after configuring your AWS credentials.

  5. Verify: After the pipeline has been running for a short time, check your S3 bucket. You will see new files appearing, organized by date. Each file will contain a batch of 10 log messages.

    aws s3 ls s3://${S3_BUCKET_NAME}/logs/ --recursive

You have now built the complete tutorial pipeline: it generates data, adds lineage, restructures it, and exports it to cloud storage.